gms | German Medical Science

Learning to diagnose patients is among the most difficult challenges medical students face. The idea that group interaction as an instructional practice boosts individual learning plays a crucial role in medical education and beyond. Recent evidence form an experimental study suggests that group-interaction may also support the diagnostic process: diagnosing (virtual) patients in pairs as opposed to individually results in higher diagnostic accuracy. Whether or not this finding translates to the workplace is currently unknown. Consequently, this study thus aims to answer two questions: Can teamwork during the diagnostic process, first, increase the quality of the diagnostic process and working diagnosis (RQ1) and, second, have long-lasting benefits for students in learning diagnosis-relevant concepts (RQ2)?

To investigate these two questions, we will conduct a randomized experimental field study at the emergency room at Inselspital Bern. Students in their clerkships will be assigned to one of two conditions: In the intervention condition, students will obtain a patient's history and physical examination in pairs. In the control condition, students will see patients individually. After an encounter, students in both conditions will be asked to provide a working diagnosis and then report to a supervising physician to decide on further diagnostic testing and treatment. Both students and patients will be asked to consent before being included in the study and only patients triaged to be non-vitally threatened will be evaluated for inclusion. A total number of 1344 patient diagnoses by a total of 48 different students will be recorded. In order to answer RQ1, students' working diagnoses will be compared against the patients' discharge diagnoses. We hypothesize that students in the intervention condition have a higher likelihood for a match between working diagnosis and the final discharge diagnosis compared to students in the control condition. Moreover, measures of the diagnostic process (including the appropriateness of selected tests, accuracy of interpretation and treatment suggestions) will be compared between conditions. We expect that students in the intervention condition make fewer errors during the diagnostic process. In order to address RQ2, we will assess students' knowledge gain on diagnosis-relevant concepts and procedures before and after their eight week clerkships and compare their performances across the two conditions. We expect students from the intervention group to show a higher learning curve.